Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting

Disheng Liu, Tuo Liang, Chaoda Song, Yu Yin

Recent generative models can produce high-quality synthetic images, offering scalable training training data for data-hungry models. Existing approaches to exploiting this potential typically involve 1) training or fine-tuning generators, or 2) using lightweight post-hoc adaptation like prompt engineering or inference-time guidance, making them generator-specific and expertise-intensive. We study a complementary question: given a fixed pool of generated images, can downstream utility be improved purely by selecting an informative subset? The answer is yes. We show that effective selection must counter a structural bias of modern generators: they tend to over-produce canonical modes of each class while underrepresenting intra-class variation. Building on this insight, we split each real class into a canonical Homogeneous (HO) subset and a non-redundant Heterogeneous (HE) subset, then score synthetic images by a fidelity-diversity criterion that rewards semantic alignment while penalizing canonical redundancy. The method is generator-agnostic and requires no retraining. Across multiple benchmarks, it consistently outperforms state-of-the-art data selection baselines and matches the real-data performance with up to 40% fewer synthetic samples. The same criterion remains effective when applied on top of stronger task-tuned generators, with gains on both classification and segmentation tasks. Post-generation selection is therefore not a substitute for better generators, but a complementary mechanism for improving the utility of synthetic data.